Method and system to compute hemodynamic parameters
Methods and systems are described herein for hemodynamic parameter estimation. In certain embodiments, a set of perfusion data is acquired for a region of interest using an imaging system. An artery signal is obtained from the set of perfusion data. A tissue signal is obtained from the set of perfusion data. The artery signal and the tissue signal are provided as inputs to one or more neural networks to determine one or more hemodynamic parameters for the region of interest. The one or more neural networks are trained using one or more synthetic data.
1 . A method, comprising:
acquiring a set of perfusion data for a region of interest using an imaging system;
obtaining an artery signal from the set of perfusion data;
obtaining a tissue signal from the set of perfusion data; and
providing the artery signal and the tissue signal to serve as inputs to one or more neural networks to determine one or more hemodynamic parameters for the region of interest, wherein the one or more neural networks are trained using modified synthetic tissue data generated based on one or more synthetic data modified or combined with one or more clinical perfusion data, wherein training the one or more neural networks comprises:
generating a set of synthetic residual impulse functions for the region of interest based on a defined ground truth model;
obtaining a training artery signal from a training set of perfusion data;
generating a modified synthetic tissue signal based on the set of synthetic residual impulse functions and the training artery signal; and
training the one or more neural networks using a signal generated using the modified synthetic tissue signal and the training artery signal, wherein a loss is used as a bias for training the one or more neural networks, and wherein the loss is determined based on a comparison of an estimated residual impulse function output from the one or more neural networks and a first set of parameters derived from the estimated residual impulse function with the set of synthetic residual impulse functions and a second set of parameters derived from the set of synthetic residual impulse functions.
2 . The method of claim 1 , wherein the set of perfusion data comprises at least one of computed tomography (CT) perfusion data, magnetic resonance imaging (MRI) perfusion data, positron emission tomography (PET) perfusion data, single photon emission computed tomography (SPECT) data, or ultrasound imaging data.
3 . The method of claim 1 , wherein the one or more synthetic data are generated based on the defined ground truth model.
4 . The method of claim 1 , wherein the tissue signal is a convolution of the artery signal and a residual impulse function of the region of interest, and wherein the one or more hemodynamic parameters are determined from the residual impulse function.
5 . The method of claim 1 , comprising correcting non-idealities in the set of perfusion data based on output from the one or more neural networks.
6 . The method of claim 1 , wherein the one or more hemodynamic parameters comprise at least one of a blood flow (BF), a blood volume (BV), a mean transit time (MTT), or a time to maximum (TMAX).
7 . A system comprising:
one or more processors; and
memory, accessible by the one or more processors, the memory storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
receiving a set of perfusion data acquired using an imaging system to image a region of interest;
obtaining an artery signal from the set of perfusion data;
obtaining a tissue signal from the set of perfusion data; and
providing the artery signal and the tissue signal to serve as inputs to one or more neural networks to determine one or more hemodynamic parameters for the region of interest, wherein the one or more neural networks are trained using modified synthetic tissue data generated based on one or more synthetic data modified or combined with one or more clinical perfusion data, wherein training the one or more neural networks comprises:
generating a set of synthetic residual impulse functions for the region of interest based on a defined ground truth model;
obtaining a training artery signal from a training set of perfusion data;
generating a modified synthetic tissue signal based on the set of synthetic residual impulse functions and the training artery signal; and
training the one or more neural networks using a signal generated using the modified synthetic tissue signal and the training artery signal, wherein a regularization is used in a bias used for the training of the one or more neural networks, and wherein the regularization is associated with characteristics of an estimated residual impulse function output from the one or more neural networks.
8 . The system of claim 7 , wherein the set of perfusion data comprises at least one of computed tomography (CT) perfusion data, magnetic resonance imaging (MRI) perfusion data, positron emission tomography (PET) perfusion data, single photon emission computed tomography (SPECT) data, or ultrasound imaging data.
9 . The system of claim 7 , wherein the one or more synthetic data are generated based on the defined ground truth model.
10 . The system of claim 7 , wherein the tissue signal is a convolution of the artery signal and a residual impulse function of the region of interest, and wherein the one or more hemodynamic parameters are determined from the residual impulse function.
11 . The system of claim 7 , wherein the one or more neural networks are trained to correct image non-idealities in the set of perfusion data.
12 . The system of claim 7 , wherein the one or more hemodynamic parameters comprise at least one of a blood flow (BF), a blood volume (BV), a mean transit time (MTT), or a time to maximum (TMAX).
13 . A method for training one or more neural networks, comprising:
generating a set of synthetic residual impulse functions for a region of interest based on a defined ground truth model;
obtaining an artery signal from a set of perfusion data;
generating a modified synthetic tissue signal based on the set of synthetic residual impulse functions and the artery signal; and
training the one or more neural networks using a signal generated using the modified synthetic tissue signal and the artery signal, wherein a loss is used as a bias for the training of the one or more neural networks and wherein the loss is determined based on a comparison of an estimated residual impulse function output from the one or more neural networks and a first set of parameters derived from the estimated residual impulse function with the set of synthetic residual impulse functions and a second set of parameters derived from the set of synthetic residual impulse functions.
14 . The method of claim 13 , wherein the modified synthetic tissue signal comprises a perturbation related to perturbating of the perfusion data.
15 . The method of claim 14 , wherein the perturbation is associated with registration errors or with acquisition errors.
16 . The method of claim 13 , wherein the set of perfusion data comprises at least one of computed tomography (CT) perfusion data, magnetic resonance imaging (MRI) perfusion data, positron emission tomography (PET) perfusion data, single photon emission computed tomography (SPECT) data, or ultrasound imaging data.
17 . The method of claim 13 , wherein the first set of parameters comprises a first set of hemodynamic parameters and the second set of parameters comprises a second set of hemodynamic parameters.
18 . The method of claim 13 , wherein a regularization is used in a bias used for the training of the one or more neural networks, and wherein the regularization is associated with characteristics of the estimated residual impulse function.
19 . The method of claim 18 , wherein the regularization comprises a second order of differentiation of the estimated residual impulse function.